Data Mining and Socio-Economic Crisis

From Adaptive Cycle
Jump to: navigation, search

Contributors

Bjorn Burscher, NN

Many crises in techno-socio-economic-environmental systems are caused by random coincidences or overcritical perturbations, which trigger cascade failures (also known as domino or avalanche effects) in such a way that the impact of random local events or perturbations becomes systemic in size. The sensitivity of the system often results from the occurence of instabilities, which create fertile ground for so called regime shifts. Such regime shifts happen at so called critical points (“tipping points”). Interestingly, when a system gets close to a tipping point, it is often characterized by[1]

  1. slow relaxation (recovery) from perturbations,
  2. increasing auto-correlations, and
  3. critical fluctuations (a large variability).


Therefore, these features can serve as advance warning signs. Since they can be determined from empirical data, massive data mining will be able to increase the level of awareness of upcoming crises and to trigger early preparations in order to avoid or mitigate them.

Data Mining can have the following purposes:

  • It can reduce serious gaps in our knowledge and understanding of techno-social- economic-environmental systems.
  • Crises Observatories (analyzing and mapping financial and economic stability, conflicts, the spreading of diseases...) could predict crises or identify systemic weaknesses, and help to avoid or mitigate impacts of crises.
  • Real-time sensing and data collection (“reality mining” of weather data, environmental data, cooperativeness, compliance, trust, ...) could reduce mistakes and delays in decision making, which often cause an inaccurate or unstable system management.

Advanced Data Mining techniques have not been extensively applied to anticipate and fight systemic crises in the past. They certainly promise better solutions for the future, supporting crises containment and the detection of feedback loops and possible cascading effects, before they cause wide-spread damage. They should be an integrative part of new ICT concepts for an adaptive risk management, facilitating and supporting a better disaster preparedness and response management.

Socio-economic Data Mining is interesting for governmental, non-profit and commercial organization. It can help them to observe various internal as well as external variables. For example, it can facilitate the prediction of markets to estimate future economic developments, outcomes of elections, fashions, the spreading of diseases, and socio-economic trends. These areas are now becoming an own business branch, complementing classical consultancy, offering services like: real-time measurement of actual user activity, identification of trendsetters, opinion leaders, and innovators in social networks, trend prediction, trend tracking, etc.

References

  1. Helbing, D., & Balietti, S. (2011). From social data mining to forecasting socio-economic crises. The European Physical Journal, 195, 3-68.

Contributors

Bjorn Burscher, NN